 Welcome back, everyone, to live CUBE coverage here in San Francisco, Google Next 23. This is theCUBE's two and a half days of coverage. I'm John Furrier, your host with Rob Streche, lead analyst with theCUBE. Collective, we have Lisa Martin here, as well as Dustin Kirkland, our CUBE contributor analyst, and we're here with Keri Tharp, who's the VP of Strategic Industries at Google Cloud, practitioner, developer in her life. Now she's at Google, bringing the goodness of Google Cloud, which is basically the AI wave hitting full throttle on the industry. It's a gift that just fell out of the sky for all the industries to leverage. Keri, thanks for coming into on theCUBE. Of course, excited to be here. We've been saying that AI's been a gift for many, because if you're in cloud and you're doing next gen kind of things like cloud scale and data, it's a great tailwind boost because if you're already in kind of with machine learning and doing good data work, the generative AI with these foundation models, there's actually a new kind of way to continue that momentum and scale faster and people are moving the ball down the field big time in new functionality. We've been reporting early about the extensions and beddings with vector databases. Some cooler new things are happening that are specific that add value to that. And it's been said on theCUBE and in the industry, every industry's impacted. That's your area. So you probably agree with that thesis. Tell us what your vision is. How do you see the market rolling out? So very exciting. I make it akin to the beginning of e-commerce in retail. It's a very disruptive force that we see actually reshaping the value chain in every industry. So the common process and steps that everybody's been working on for years using traditional AI really in an optimized, get more predictive, get a little bit better. Generative now changes the scale and scope of how you can reshape those processes. So whether that's customer facing, improving employee productivity, driving process automation. So things that were kind of a dream in CIO, CTOs, minds, CEOs of how do I get more revenue, less cost. You now have a whole new toolkit to go after that. Your role at Google is not so much product, but you're facing the customer because you've been a practitioner, been there, done that. So you're more of a trusted, credible advisor slash consultant, I don't know what word to use, but to help the customers figure out how to do this, AI stuff, hang with Google Cloud. That get that right? Absolutely, it's how do you put the one plus one to get three from product, but apply it to a business use case and problem? So right now it's pretty clear from the dorm room, boardroom to the dorm room is AI madness. Hope we've been talking with your colleagues around the developer frothiness and the excitement around the coding, open source, long tail power law, models are coming out, you have proprietary, it's just if you're under the age of 30 and you're technical or entrepreneurial, you're doing AI stuff, it's really kind of cool. So bottoms up is going great. Top down is take that hill, put everything AI and everything, the directive. Now the implement is like, okay boss, how do we do that? That's been challenging and we've been finding out. The innovation strategy, we see that clearly, but compliance, legal. There's a lot of architectural things that practitioners have to kind of rethink or look at to kind of make it happen. That's kind of like the current blocker. The only solution there is to figure it out, right? So that's what you're trying to do. So how do you see customers figuring out how to look at the bets they got to make? Architectural decisions that are on the table, what changes for them in their current landscape and their journey for transformation? Yeah, so first we surveyed our top customers 89% top priority to implement Gen AI, but the prerequisite to that is 65% understand they have to start with the business case and kind of stack rank those priority use cases. This is where you're seeing people create a range of risk profile because if there's regulatory compliance, there's customer risk, brand and IP at risk, you're seeing those use cases kind of be shifted to the back and others brought forward. Thinking about it from an architecture standpoint, this is kind of data everywhere. Limitless data, data used to not be able to connect and create relationships with, now you're going to use it in these immersive use cases. And so having to step back from that entire data foundation you've been building and understand when I bring these data sources together, when I have to think about HIPAA compliance, whatever it might be, creating a new practice inside these organizations to really understand the dream of the use case has now been unlocked, but there's actually 15 steps that you have before that to really bring these use cases to bear, whether it's in a patient environment, a customer environment, even in financial environment. So it's rethinking kind of those skill sets as well as the architecture itself. And you're hearing us talk about it's a platform. You have to think durably about not a single use case, not a single data source, and how you architect that to create a platform for growth. So architecture is really the big deal. Absolutely. And how does that differ in what industries are you covering? Because saying industries means a lot of different things to a lot of different companies, I guess you could say. Yeah, so when we think about industries, it's kind of the core sector. So you see clustering and consumer industries that are really focused on things like customer experience, changing the flow of digital traffic online, regulated industries that all have kind of these key considerations. So that's healthcare, financial services, telco, et cetera. And they're kind of really thinking about how they bifurcate all those use cases. But really any business that all of us interact with as a consumer or patient, games, media, entertainment, they all have specific uses. And so we're building products and solutions and models against those specific uses. Rob and I were talking at the top of the show during our keynote analysis, during our analyst angle session, we were kind of analyzing it. Not a lot of critical analysis this year with Google. So you guys really kind of moved the ball down the field, looking great. We said if you guys can pull off the trifecta, then that would be a game changer for the company. First move is developers. Google's got a great history of developers, well-documented in Google proper, Google Cloud as well, open source, goodness there. Okay, win the development. That's a nice shot at Amazon because they have developers too. And now the new younger demographic shifting with AI, that's an opportunity. Two, solutions. Not going down and going about services and higher levels of just build some solutions and three, ecosystem. So we're seeing the ecosystem here. So I want to ask about the ecosystem because as you go to your customers, they now have an ecosystem partner network to pull from. MongoDB's got Atlas database, really great product. You got GSIs out there that can maybe lead the way. How do you see that going? And how do you interact with the customer as they bring the ecosystem to bear? Because now it's more power to the table with the partners. This is one of the most important areas when we think about the future of AI. Everybody who's been around the block before knows that in traditional AI, it was a GSI, ISV, kind of partner ecosystem. You aren't going to build everything yourself. There's complex stacks in all these industries. In many cases, a lot of legacy elements that are kind of dragging with you to perform these use cases. So that's been a big focus of ours to ensure that we have this open architecture bringing in that ecosystem so that not everybody's going to build out big data and analytics and prompt engineering team and really kind of explode that out and they need to be working with the ecosystem to be able to deliver business value. This is the top conversation we're hearing at board levels. Even if they have no idea what it actually means or asking the CTOs and CIOs where are you at on this? And so making sure, regardless of the scale or sophistication of a company, they have that ecosystem to pull on. So that's important to you guys when you're out in the field with customers. What's their response? Does they have most of these vendors in there? Yeah, so it's usually excitement because there's a comfort factor if you're working with certain partners already, you're familiar with that. It's an integrated tier stack. It might be a part of your e-commerce experience today. That sounds like faster path to value. And so that's what's important to our customers. So where can you work with these other partners? It also creates kind of the safety net where you have somebody that's really kind of pounding against these models doing all the testing and bringing that ready-made, purpose-built solution to your business. Lisa O'Malley was in earlier, she was talking about industries and having domain expertise in those areas. One of the IP advantages we're seeing with data is that in certain domains, so property is the data. And as companies have that rich domain expertise, it's in data, they also got to make it widely available. So the question comes up is, how do they figure out, architect that? And so I'd love to get your perspective of, when you look at real-world problems that are opportunities, a company that refactors around AI could actually go out and solve problems with that domain. They're not asking the Google main model or some proprietary model, they have their own data. So like the data in the domain of the vertical or the industry is proprietary intellectual property now. Now that's a little, I mean, there's kind of been IP, oh yeah, data on the balance, you've heard those conversations in the past decade, but never like this, never like this. What's the role of that specialized data? That is the secret sauce. So every model that you talk about, it's only as good as what it knows. And so when you look at all the enterprise data, the knowledge of brands, the knowledge of customers, that's where these really kind of breakthrough experiences come to bear. You've heard industries talk about monetizing their data for a long time. This is a case of really them accessing their own data for a higher business purpose than they've ever been able to before. This is not a report that you're going to look at and say, hey, maybe we do this a little bit different tomorrow. This is taking all of that pent up data and the power behind it and using it to drive the P&L. It would seem that a lot of your conversations have to start with security, data leakage, how do I prevent privacy concerns? And also, how do I then ground my models in that data? Is that where you start with the, after the business case, I'm sure. That is the next top conversation and that's part of what we've been talking about here today, that your proprietary data is not used to train the models. Whether you're interacting with Duet or Vertex AI, whatever you're bringing, brand images, customer data, that's yours operating up against effectively a version of the model. And that's the secret sauce. So we've had customer announcements with PriceLine, for example, that are creating new generative experiences for travel planning. They want their view of travel and their view of the customer to stay proprietary to them. And so us ensuring that through the platform becomes this momentum driver because it's the top question that we have from executives. Kerry, you mentioned earlier in the interview that AI reminds you e-commerce. Obviously you had probably some e-commerce background in your day. We've been talking a lot about how it's like the web early days e-commerce where the obvious future cannot be debated. Everyone is all in, it's obvious it's going to happen. I think that's true as well as it's early and not the functionalities still not great getting better every day. And the online usage population of users in that era was online population of web users. And so now the online usage of AI production system, we see that growing too. So we see this happening. So there's no real debate there. The question is, what happens next? Because like the iPhone or phone and Android iPhone debate, we're kind of in an AI mode now where you got to make bets. You got to pick your bet, trusted partner. Because if you've got a vector embedding database and you start going changing embeddings over here, you don't have to get all nerdy, but like that could matter. So how do you square that up in your mind with your experience with folks watching? Because I think the number one question I get is, what do I bet on? Which course is out there? Where am I going to have headroom? I don't want to foreclose the future opportunity. I want to have business value. How do you look at that? And it's a really great analogy that you brought that up. I want to bring back that out. Yeah. To me, it's one of those, when you get in a business, you can get kind of blinders on, you know what you know. And it's kind of what we say in industry, you're talking to yourself about that experience. I think what's happening in generative, it allows us to break the structure, frankly, of the internet. So back in the 90s, we trained everybody, search and nav, search and nav, search and nav. And people have stood up chatbots and digital experiences in past years, but the fidelity wasn't there, so the user traffic didn't follow. So as everybody's thinking about whether it's an internal use case or out there on the web, I think it reshapes the flow of digital experience. So I think even just the structure of how we think about websites changes. That's multiple years out, but it's, in some of these industries like retail, travel, et cetera, you've gotten used to the box that you're in and this will break the box. So when you're thinking about value, you've got those near-term use cases run after those as fast as you can. But I think it's a whole new experimentation phase. It's going to be a lot of failed test and learn where nobody likes that experience, but eventually something's going to break through. And it's often the scale players in each of these industries that set a new bar for what that digital experience is. And so right now everybody should be watching. We're going to see hundreds of failures, but then you're going to start to see those breakouts that become the future of our digital interaction. Yeah, new brands will emerge from entrepreneurial circles. You're going to see the next Google possibly, the next Airbnb pop out or something. This disrupts industries and creates new business models. So that's why I say it's the most exciting time since the beginning of the internet. So Kerry, would you agree then if you said the common thread between that comment about the web and commerce and today is that the people keep pushing the experience value proposition? There's friction. We all know in every experience we have, whether it's going to your doctor, whether it's ordering a product online, there is still friction in all those experiences today based on technology and human limitations. And so as long as you continue to follow that trail, that's where you need to be going. You know, making things easy, simple and reducing the steps it takes to do something is a great business model, great value. But that's kind of the name of it, it's so simple. That is all the disruptors. When you think of like the Warby Parkers, the Ubers, et cetera, it all started with, there's this challenge that I could solve differently than I used to be able to. And AI is going to do that faster. Final question for me, Rob's probably got another question too. I've been taking all the air time here. Is, you know, with AI, speed to value is probably unprecedented. You know, product market fit, you heard that term many times in the tech industry. So business value can be captured faster. That's also going to disrupt the teams. So the skills, the teams, the time to value is accelerated and maybe have more impact. What is your kind of, being a kind of a historian because you've been in the movie before, how do you see that coming and how should people be prepared for that? Or do you agree with that and what do you see that rolling out? Absolutely agree with it. There's a lot of people in change management. Processes are changing, but even if you think about traditional AI just a year or two ago, I've sat in rooms where people debated the answer from the model and they were worried about things like explainable AI. Why is it this answer versus what the humans came up with? And this really just kind of explodes that experience into dimensions that the human brain can't solve and put in a box. And so working with the teams to understand what are the new test standards? How do you look at getting something to production when it's a customer experience? What's good enough that you're going to keep testing your way through? So I think it's going to change a lot of those processes and timelines. Do you think it'll change the way that organizations measure their developers and development? Like, so like your Dora report will come out pretty soon. I'm suspecting any day now. He's a big fan. I'm a big fan of that. I read it every year and it's, you know, the geeky stuff underneath there, but being developer background as well. I mean, but how do you see that and what are you telling the organizations for them to prepare for that? It's kind of expanding your skillset because the metrics in OKR has changed because as you can see in code generation and duet kind of sitting by your side, your ability to create apps, to debug, all of that has fundamentally changed. And so what used to be the performance bar has now changed. Those that rise to the occasion and use these tools will accelerate, create new innovation and others. You know, it's like all things. There are folks that will kind of stick to the way they used to do it. And you'll kind of see that create disruption in the knowledge workers, the developers themselves. So just like industry, the workers will kind of have that disruptive force in their own experience. Awesome. Carrie, thank you for coming on theCUBE and sharing the data with us and exciting. I think the industries are going to be disrupted either by themselves or by someone coming in. And I just such an exciting thing. You know, we've been riffing on theCUBE about, you know, how old we are in the industry and AI is like a fountain of youth. Everyone I talk to that has systems background or been in practitioner roles. When they see the AI stuff going on, they're like, oh my God, this is like a jolt of energy because it's so intoxicating at so many intellectual levels of the process change. You make some of those. It's just, it really is legit next level. Yep. That's my favorite kind of parting thought. If you don't disrupt yourself, someone else will. So you pick your path. Carrie, it's our VP of strategic industry at Google Cloud. She's out there talking to the customers, understanding how to shape them into the cloud era of AI and say AI clouds theCUBE, bringing you all the data live. John Furrier here with Rob Stretching. We'll be right back with our next segment after this short break.